申请学校时，教授通常会要求先提交一份research statement，以此来考察对方的语言功底，以及研究兴趣、思路是否和自己的实验室很匹配。（tips：去看一看所申请的教授的个人主页，说不定他就有自己的research statement，可以参考着写）

Research Overview

总体介绍自己参与过的研究方向，感兴趣的技术等。(篇幅无所谓)

My primary research interests lie in machine learning and bioinformatics. Machine learning is an area of computer science that studies and develops methods for identifying and exploiting interesting regularities in data. Bioinformatics is the field that develops computational methods for attacking problems in biological domains. I am particularly interested in systems that learn from complex and diverse data sources. There is a strong need for such systems since technological advances have led to increases in the types and amounts of digital data while automated methods for the challenging problem of extracting knowledge from these data have lagged. In addition to the proliferation of digital text, audio, and images, high-throughput technologies in biology, medicine and other fields generate large amounts of data that are not amenable to manual analysis. Since these technologies often provide first-of-their-kind glimpses into fundamental processes, there is a keen interest in computational methods for interpretation of the resulting data.

I have a special interest in applications of probabilistic methods to bioinformatics problems. My interest in bioinformatics stems from two sources. First, bioinformatics problems have qualities that make them wellsuited for the study of machine learning and to the application of probabilistic methods. These characteristics include inherent uncertainty, complex dependencies, diverse evidence sources, and the availability of labeled training data and domain knowledge. Second, bioinformatics problems typically involve challenging and often fundamental open issues in biology and medicine, and working towards solving them is both gratifying and substantial.

My research to date reflects these cross-disciplinary interests. Its unifying theme concerns the development of machine learning methods for prediction of biologically important elements. The style of my research is to push the state of the art in learning algorithms by working on hard computational problems in biology. My research has made both biological and computational contributions. In biology, my predictions have been used by researchers around the world. In computer science, my models and algorithms have proved to be of broad interest and have been published in leading conferences and journals.

Research to Date

这里介绍自己的研究经历，如果接触过多个不同方向的项目，可以分段分别来介绍。对于自己发表过论文的项目，需要简要介绍背景意义和成果。tips：可以添加引用，并在statement最后加上reference参考文献，这样可以体现自己所发表论文的level。

Bioinformatics Contributions: The focus of my doctoral research is on developing methods for the prediction of key DNA sequence elements involved in the process of transcription in bacteria. Since genes are frequently “turned on” and “turned off” at the level of transcription, knowledge of these elements is useful to the study of gene regulation. Indeed, my predictions for the widely studied E. coli genome have been used by researchers in this area. Additionally, my predictions provide insight into the function of genes and the evolutionary relationships among organisms. The basic task I address is the following: starting with example instances of various types of elements, first use machine learning methods to induce models that capture important characteristics of the training instances, then apply these models to predictively identify elements within sequences of interest.

The primary contribution in this area has been the development of a model to predict the transcription units of a bacterial genome. My approach has several advantages relative to related work: (i) it predicts the complete extent of transcription units, (ii) it makes genome-wide predictions for multiple types of elements while enforcing a consistency among all predictions, and (iii) it simultaneously considers multiple sources of evidence. This research makes me well positioned to make further advances in bioinformatics. In particular, I believe methods that coherently combine data from many different sources will become more important as more high-throughput technologies for interrogating the states of cells emerge.

Contributions II: ……

Future Research Directions

对未来方向迷茫怎么办？写出来没深度怎么办？不用急，找几篇自己领域感兴趣的论文，他们的future work就是你未来研究的方向。